Micro-engineering is a rapidly growing field which is liable to impact on many applications over the coming years. Three-dimensional micro-engineered devices and systems involving silicon planar technology can be produced with features from one to a few hundred microns having tolerances in micron or as small as submicron level. Most of the current micro-engineering technologies are evolved from the adaptation of thin films, photolithographic and etching technologies generally applied to silicon wafers on which silicon monoxide, silicon dioxide, silicon nitride and the like thin films are deposited and etched thereafter yielding planar configurations.
Advances have been made to study chemical processes based on microfluidic systems technology using planar structures on silicon chips. Predictive flow modeling has also been developed to realize the benefits from the microfluidic technology.
The performance of chemical processing is strictly governed by the mass transport and sometimes thermal transport properties of the system. It is therefore essential to understand how miniaturization affects the chemical processes. Laminar flow of an ideal fluid in a tube or channel is well characterized. Pfahler et al (J. Pfahler, J. Harley, H. Bau and J. Zemel; sensors and Actuators; Vol. 21-23 (1990); page 431-434) have demonstrated in an experiment using channels of various geometry on silicon wafers that there is an agreement between experiment and theory concluding that the conventional Hagen-Poiseuille equation is obeyed down to a scale of few microns. Laminar volume flow per unit time, Q.sub.f, of an ideal fluid in a circular pipe is described by the Hagen-Poiseuille equation: ##EQU1## where .mu..sub.f and r are the fluid viscosity and tube radius respectively, dp/dx is the pressure gradient along the x-direction of flow.
As the channel widths are reduced, the fluid flow becomes more laminar which provides control over the distribution of material and that dictates that fluid mixing is achieved by diffusion or other molecular migration processes rather than by turbulence. This problem of mixing can be solved by commercially available software packages on computational fluid dynamics. A measure of degree of mixing, F, can be estimated from the expression F=Dt/1.sup.2, where D is a reactant diffusion constant, t is contact time allowed for mixing and 1 is distance across a reactant stream. Quantitatively, mixing may be defined as substantial to nearly complete for F values from 0.1 to 1. Typically, near complete mixing of two fluids in 1 second corresponds to channel widths of 100 .mu.m.
Similarly, problems exist with respect to heat transfer in micro-channels under laminar flow conditions. Understanding of this laminar heat flow process can be useful in designing and building micro heat exchangers and chemical micro-reactors.
The current planar silicon technologies are inadequate for the fabrication of an integrated and self-contained catalytic reaction and micro-filtering arrangement.